Which of the following is the GREATEST challenge facing IS auditors evaluating the explainability of generative AI models?

The correct answer is D. Algorithms changing as AI continues to learn. Generative AI models that continuously learn update their internal weights and decision logic over time. This means the model that produced an output last month may operate differently today, making it extremely difficult to audit or explain past decisions with current model…

AI Audit Planning and Execution

Question

Options

ADifferences of opinion regarding model types
BDifficulties in preventing the input of biased data
CPerformance issues due to excessive computation
DAlgorithms changing as AI continues to learn

How the community answered

(16 responses)

A
6% (1)
B
25% (4)
C
6% (1)
D
63% (10)

Explanation

Generative AI models that continuously learn update their internal weights and decision logic over time. This means the model that produced an output last month may operate differently today, making it extremely difficult to audit or explain past decisions with current model state - and equally hard to predict future behavior. This dynamic nature is the core explainability challenge. Differences of opinion (A) on model types are a professional, not a technical, challenge. Preventing biased data input (B) is a data governance issue, not an explainability issue. Computational performance (C) affects speed and cost, not whether decisions can be explained.

Topics

#Explainable AI#AI Auditing Challenges#Model Drift#Generative AI

Community Discussion

No community discussion yet for this question.

Full AAIA Practice